基于差分拉曼光谱-化学计量学集成架构的塑料包装泡沫分类研究

姜红; 张贤程

doi:10.19554/j.cnki.1001-3563.2026.07.022

PDF(1494 KB)

包装工程（技术栏目） ›› 2026, Vol. 47 ›› Issue (7) : 186-192. DOI: 10.19554/j.cnki.1001-3563.2026.07.022

自动化与智能化技术

基于差分拉曼光谱-化学计量学集成架构的塑料包装泡沫分类研究

姜红^1,2,*, 张贤程³

作者信息 +

Research on Plastic Packaging Foam Classification Based on a Differential Raman Spectroscopy-chemometrics Integrated Framework

JIANG Hong^1,2,*, ZHANG Xiancheng³

Author information +

文章历史 +

摘要

目的建立一种高效、无损的塑料包装泡沫系统化分类方法。方法利用差分拉曼光谱采集63个塑料包装泡沫样品的光谱数据,采用主成分分析和线性判别进行特征降维与提取;构建K近邻、支持向量机、轻量级梯度提升机与极端随机森林的集成分类模型;基于理论可分上限评估数据的分类潜力,并采用多层感知器对模型的泛化能力进行交叉验证。结果该架构依据特征峰将样品分为五大类。经PCA-LDA降维后,集成模型在测试集上分类准确率达93.33%;MLP验证的训练集与测试集准确率分别为95.45%与89.47%。结论构建的集成分类架构融合了理论评估与多模型集成策略,实现了分类性能的极限逼近。该方法为快速鉴别塑料包装泡沫提供了一套可靠完整的解决方案。

Abstract

The work aims to establish an efficient, non-destructive systematic classification method for plastic packaging foams. Spectral data from 63 plastic packaging foam samples were acquired by differential Raman spectroscopy. Feature dimension reduction and extraction were performed via principal component analysis and linear discriminant analysis. An ensemble classification model was constructed incorporating k-Nearest Neighbors (k-NN), Support Vector Machines (SVM), Lightweight Gradient Boosting Machine (LightGBM), and Extreme Random Forest (ERT). The classification potential of the data was evaluated based on the theoretical separability threshold. The model's generalization capability was independently validated using a Multi-Layer Perceptron (MLP). Results showed that the architecture classified samples into five major categories based on characteristic peaks. After PCA-LDA dimension reduction, the ensemble model achieved a classification accuracy of 93.33% on the test set. The MLP-validated training and test set accuracy rates were 95.45% and 89.47%, respectively. In conclusion, the constructed ensemble classification architecture integrates theoretical evaluation with multi-model ensemble strategies, achieving near-limit classification performance. This method provides a reliable and comprehensive solution for rapid identification of plastic packaging foams.

导出引用

姜红, 张贤程. 基于差分拉曼光谱-化学计量学集成架构的塑料包装泡沫分类研究[J]. 包装工程. 2026, 47(7): 186-192 https://doi.org/10.19554/j.cnki.1001-3563.2026.07.022

JIANG Hong, ZHANG Xiancheng. Research on Plastic Packaging Foam Classification Based on a Differential Raman Spectroscopy-chemometrics Integrated Framework[J]. Packaging Engineering. 2026, 47(7): 186-192 https://doi.org/10.19554/j.cnki.1001-3563.2026.07.022

中图分类号： TB48 O657.37

参考文献

[1] 赵艳, 潘祥, 刘本刚. 可发性聚苯乙烯土工泡沫的性能及应用[J]. 中国塑料, 2021, 35(5): 97-106.
ZHAO Y, PAN X, LIU B G.Properties and Applications of Expendable Polystyrene Geofoam[J]. China Plastics, 2021, 35(5): 97-106.
[2] 王剑侠, 姜红, 张晓璐. 傅里叶变换红外光谱法检验EPS泡沫塑料的研究[J]. 上海塑料, 2017, 45(3): 27-32.
WANG J X, JIANG H, ZHANG X L.Research of Expandable Polystyrene Foamed Plastics by FTIR[J]. Shanghai Plastics, 2017, 45(3): 27-32.
[3] 姜红, 鞠晨阳, 盘永平, 等. 拉曼光谱法检验泡沫塑料的研究[J]. 上海塑料, 2019, 47(2): 48-53.
JIANG H, JU C Y, PAN Y P, et al.Research on Foam Plastics by Raman Spectroscopy[J]. Shanghai Plastics, 2019, 47(2): 48-53.
[4] 丁建明, 王鑫, 张容玲, 等. 拉曼光谱结合偏最小二乘法定量分析水中聚乙烯和聚苯乙烯两种微塑料[J]. 分析化学, 2024, 52(10): 1581-1590.
DING J M, WANG X, ZHANG R L, et al.Raman Spectroscopy Combined with Partial Least Squares for Quantitative Analysis of Two Kinds of Microplastics in Water Samples[J]. Chinese Journal of Analytical Chemistry, 2024, 52(10): 1581-1590.
[5] 余敏. GC-MS/MS法检测聚氨酯泡沫塑料中14种有机磷酸酯阻燃剂[J]. 中国塑料, 2023, 37(10): 8-14.
YU M.Determination of 14 Organophosphate Ester Flame Retardants in Polyurethane Foam through GC-MS/MS[J]. China Plastics, 2023, 37(10): 8-14.
[6] 刘彤彤, 黄登丽. 聚氨酯泡沫塑料吸附-电感耦合等离子体质谱(ICP-MS)法测定地球化学样品中的痕量金[J]. 中国无机分析化学, 2021, 11(2): 45-50.
LIU T T, HUANG D L.Adsorption and Thiourea Desorption of Trace Gold in Geochemical Samples by Foam Plastics[J]. Chinese Journal of Inorganic Analytical Chemistry, 2021, 11(2): 45-50.
[7] 刘迎丽, 牟涛涛, 陈少华. 移频激发差分拉曼光谱技术的研究进展综述[J]. 激光与光电子学进展, 2024, 61(9): 0900003.
LIU Y L, MU T T, CHEN S H.Research Progress of Shifted Excitation Raman Difference Spectroscopy[J]. Laser & Optoelectronics Progress, 2024, 61(9): 0900003.
[8] 姜红, 陈壮, 郝小辉, 等. 基于主成分分析-Fisher判别分析的食品类塑料瓶物证差分拉曼光谱分类[J]. 化学通报(中英文), 2024, 87(1): 118-121.
JIANG H, CHEN Z, HAO X H, et al.Differential Raman Spectral Inspection of Food Grade Plastic Bottles Based on Principal Component Analysis and Fisher Discriminant Analysis[J]. Chemistry, 2024, 87(1): 118-121.
[9] 周飞翔, 姜红, 胡晓光, 等. 基于差分拉曼光谱的药品包装纸盒快速检验分析[J]. 应用激光, 2024, 44(11): 119-128.
ZHOU F X, JIANG H, HU X G, et al.Rapid Inspection and Analysis of Drug Packaging Cartons Based on Differential Raman Spectroscopy[J]. Applied Laser, 2024, 44(11): 119-128.
[10] TIAN L C, JIANG H, CHEN T Z.A Rapid and Nondestructive Approach for Forensic Identification of Novel Psychoactive Substances Using Shifted-Excitation Raman Difference Spectroscopy and Machine Learning[J]. Journal of Raman Spectroscopy, 2023, 54(5): 540-550.
[11] LEE J, JIANG H.Analysis of Indole and Indazole Amides Synthetic Cannabinoids by Differential Raman Spectroscopy Based on ANN[J]. Journal of Forensic Sciences, 2022, 67(6): 2242-2252.
[12] 周飞翔, 姜红, 骆骄阳, 等. 基于差分拉曼光谱对黄芪药材的快速检验分析[J]. 实验与分析, 2024, 2(3): 50-54.
ZHOU F X, JIANG H, LUO J Y, et al.Rapid Inspection and Analysis of Drug Packaging Cartons Based on Differential Raman Spectroscopy[J]. Labor Praxis, 2024, 2(3): 50-54.
[13] 杨环瑜, 林进超, 郭佳. 支持向量机模型在中药材产地鉴别中的应用[J]. 农业开发与装备, 2025(2): 46-48.
YANG H Y, LIN J C, GUO J.Application of Support Vector Machine Model in Identification of Traditional Chinese Medicine Origin[J]. Agricultural Development & Equipments, 2025(2): 46-48.
[14] 陈和生, 孙育斌. 几种塑料的傅里叶变换拉曼光谱分析[J]. 塑料科技, 2012, 40(6): 69-72.
CHEN H S, SUN Y B.Analysis of Several Kinds of Plastics by Use of FT-Raman Specroscopy[J]. Plastics Science and Technology, 2012, 40(6): 69-72.
[15] 吴喜之. 多元统计分析——R与Python的实现[M]. 北京: 中国人民大学出版社, 2019.
WU X Z.Multivariate Statistical Analysis with R and Python[M]. Beijing: China Renmin University Press, 2019.
[16] DOGAN M, GOKSEL SARAC M, TOKER O S.Investigation of Rheological Synergistic Interactions between Hydrocolloids and Starch in Milky Cacao Beverages Model: Principal Component Analyses[J]. European Food Research and Technology, 2017, 243(6): 1031-1039.
[17] 李博, 朱莉, 姚庆宇, 等. 基于近红外光谱的草莓多品质参数通用预测模型研究[J]. 现代食品科技, 2025, 41(8): 227-236.
LI B, ZHU L, YAO Q Y, et al.General Predictive Model for Multiple Strawberry Quality Parameters Based on Near-Infrared Spectroscopy[J]. Modern Food Science & Technology, 2025, 41(8): 227-236.
[18] 韩斌, 曾志祥, 孔繁新, 等. 基于多层感知器神经网络的风机叶片覆冰预测模型研究[J]. 发电技术, 2026, 47(1): 65-74.
HAN B, ZENG Z X, KONG F X, et al.Research on Ice Accretion Prediction Model for Wind Turbine Blades Based on Multi-Layer Perceptron Neural Network[J]. Power Generation Technology, 2026, 47(1): 65-74.